Direct detection of fast neutrons is useful for the identification of special nuclear materials (SNM), including fissile materials relevant to nuclear non-proliferation, monitoring and verification. Typically, pulse-shape discrimination (PSD) is employed in organic scintillator systems due to the high sensitivity, low cost, and effectiveness in separating fast neutron events from background gamma radiation.
Scintillation and pulse-shape discrimination capability were first demonstrated in organic molecular crystals dating back to the early 1950's. Since then, a variety of different organic liquid scintillators and scintillating plastics have been developed for use in scalable, low-cost particle detection and discrimination systems. In spite of these significant advances, however, molecular crystals remain among the best materials for high fidelity measurements and fundamental energy loss studies, due to their ordered crystallographic structures and superior pulse-shape discrimination performance.
Pulse-shape discrimination in organic scintillators depends upon differences in the specific energy loss (dE/dX) of different types of ionizing particles within the scintillator material, and the corresponding effect on the relative proportion of prompt and delayed luminescence, as generated via fluorescence from singlet excited states and diffusion-controlled triplet-triplet annihilation (TTA), respectively.
Scintillators generally refer to materials which emit prompt luminescence when exposed to ionizing radiation. When excited by ionizing radiation, electrons may be freed from atoms of the scintillating material. The electrons and molecular ions recombine to form neutral states, including “singlet” (spin zero) and “triplet” (spin 1) excitation states. Singlet excited states generally refer to states in which the excited electron (spin ½) is paired with a ground state electron with opposite spin (total spin zero). Triplet excited states generally refer to states in which the excited electron is not paired with an opposite-spin ground state electron (total spin 1).
Approximately 25% of the electrons excited by ionizing radiation in a scintillating material may relax to singlet excited states, while 75% of the excited electrons may go to a triplet state. As the excited electrons relax to a ground state, they emit luminescence, but the characteristic emission time is not the same for luminescence generated from singlet and triplet states. In addition, the relative proportion of luminescence originating from singlet and triplet states is not the same for all particles. Neutron interactions, for example, are observed via elastic scattering from protons, which tend to generate relatively less luminescence from singlet versus triplet states, while gammas are observed via electron scattering and pair production, which tend to generate relatively more luminescence from singlet versus triplet states.
FIG. 1 is a schematic illustration of representative transitions experienced by excited electrons. Excited electrons occupying singlet (S1) states 102 may generally freely relax to a ground (S0) state 103, emitting luminescence photons 110 of energy hν, where h is the Planck constant and ν is the frequency. Luminescence 110 generated based on relaxing singlet states is generally considered “fast” luminescence, occurring on a time scale of the order of nanoseconds, based on the singlet state excited electron making a direct radiative transition to the ground state in the scintillating material.
Excited electrons in triplet (T1) states 112 and 114 may not freely relax to a ground state, because transitions from a triplet (spin one) states to ground (spin zero or singlet) states are “spin-forbidden,” based on symmetry considerations, while transitions from excited singlet (S1) states to singlet ground states (S0) are “spin-allowed.” Pairs of excited electrons in triplet (T1) states 112 and 114 may however combine with one another via diffusive processes, generating an excited electron in a singlet (S1) state 116 and another electron in a ground (S0) state 118.
This diffusive process has a relatively longer time scale, for example on the order of hundreds of nanoseconds, as described below. The excited electron in the singlet (S1) state 116 can then relax to the ground (S0) state 120 via an allowed radiative transition, generating a delayed or “slow” luminescence photon 122 with energy hv. The uncombined triplet (T1) states tend to produce little or no direct or prompt luminescence, due to the lower probability of non-radiative transitions, and may instead result in much slower phosphorescence emission, with a time scale on the order of milliseconds or even longer. Such emissions are typically not significant in radiation measurements due to the low phosphorescence quantum yields of organic scintillators.
The mechanisms that control prompt emission from singlet states are understood within the framework of Förster dipole-dipole interactions, but less is known about the relative proportion and kinetics of triplet-derived (delayed) fluorescence. Studies have shown, however, that TTA efficiency depends upon the mobility and lifetime of the triplet excited states, both of which are governed by intermolecular interactions between chromophores.
The magnitude of the intermolecular interactions also depends upon the overlap between molecular orbitals (e.g., π orbital overlap). The overlap integrals are controlled by the configurational geometry and distance between adjacent molecules, which in turn are enforced by the crystallographic structure. In some organic scintillating materials delayed singlet luminescence may thus be observed, where the rate of the delayed luminescence component is determined by the rate of diffusion of the triplet states combining with one another within the scintillating material. Accordingly, delayed luminescence may exhibit a non-exponential decay profile, with a substantially longer lifetime. Typically, only a small fraction (such as two percent) of excited electrons in triplet states may undergo this recombination and relaxation to produce luminescence.
Thus, the fast (or prompt) and slow (or delayed) luminescence components can be used to discriminate between ionizing particles in scintillating systems, for example to discriminate between energetic neutrons and gamma ray photons. These are neutral particles, which must generally be converted to charged particles in order to be detected. Neutrons are typically observed based on the generation of recoil protons, whereas gammas are converted to fast electrons. Particle discrimination is possible at least in part because the relative proportion of the fast luminescence component is dependent on the energy deposited per unit distance (dE/dX) in the scintillating material, which tends to be less for electrons than protons. High ionization densities can quench the excited singlet π electrons, with non-luminescent de-excitation reducing the fast component for high dE/dX particles. This introduces non-linearity into the energy response, and results in different pulse shapes for different particle types.
FIG. 2 is a schematic illustration of the luminescence intensity generated by ionizing electrons and recoil protons, respectively. Intensity of the photon signal is indicated on the vertical axis, in arbitrary units. Time is shown on the horizontal scale, for example in nanoseconds (ns).
The luminescence or photon signal generated by a representative scattered electron is illustrated by dashed curve 202. The luminescence or photon signal generated by a recoil proton is illustrated by solid line 204. As illustrated in FIG. 2, the initial “fast” luminescence intensity 210 may vary according to dE/dx, and therefore differ between the electron and recoil proton, with the electron producing a relatively greater fast luminescence component.
In this particular example, the decay data for electron excitation 202 and proton excitation 204 have been normalized to the respective delayed luminescence signals 212. Each particle type produces a different relative proportion of fast and slow components. These effects can be used to differentiate signals from the different particle types based on the relative fast and slow components in the light signals from different particles, a technique referred to as pulse-shape discrimination (PSD).
FIG. 3 is a schematic illustration of a scintillator system 310 with scintillator 302, photodetector 304 and electronics 306, for example as configured to perform pulse-shape discrimination. In this particular embodiment, a photomultiplier tube (PMT) or multi-pixel photon counter (MPPC) type photodetector 304 is positioned to receive luminescence generated by a scintillating material 302 via a substantially direct optical coupling. Alternatively, a light pipe, optical fiber or other coupling may be used.
Photon detector 302 is configured to generate an electronic signal in response to the luminescence generated by the passage of ionizing radiation through scintillator 302. Electronics 306 is coupled to detector 304 in order to receive the electronic signal, and to process the signal in order to detect and identify sources of ionizing radiation. Electronics 306 can also be configured to discriminate between particle types based on the pulse shape of the electronic signal, as described above, based on the temporal signature of the relative fast and slow luminescence components.